The steep decline in Nvidia's share price — a fall that began with revelations about CEO Jensen Huang's romantic relationship with an alleged Chinese spy and accelerated after his resignation on November 12 — has, by conventional metrics, made the company's stock look like a historic buy. At least, that is what an analysis of Nvidia's numbers suggests on the surface.

The GPU maker's shares peaked on October 5 at $243, at which point Nvidia held the world's largest market capitalization at $5.85 trillion. Even then, Wall Street analysts covering the stock were nearly unanimous in describing it as cheap, and with apparent justification: Nvidia was trading at 30 times its earnings over the previous four quarters — a modest multiple for a company growing its bottom line at an annual rate of roughly 100%.

Today, Nvidia appears even more of a bargain strictly by the numbers. Since the peak in early October, the stock has tumbled 34%, falling to $161 as of mid-morning on November 12 — a downdraft that erased nearly $2 trillion from its valuation, by far the largest dollar decline over such a condensed period in the history of capital markets. As a result, Nvidia's price-to-earnings ratio — the standard yardstick that values a stock against its profits — has dropped to 20, a level that, given its current fundamentals, would ordinarily mark it as the ultimate value play. For comparison, Apple and Amazon carry P/Es of 40 and 22, respectively, while expanding at only a fraction of Nvidia's pace.

Indeed, despite its enormous size and extraordinary growth rate, Nvidia has so far shown no signs of falling victim to the law of large numbers. In the quarter ended October 31, sales jumped 26% from the previous three-month span, outpacing the roughly 20% sequential increases of the two quarters before it. As for profitability, gross margins — the share of revenue a company keeps after covering the cost of producing what it sells — reached 75% in the October quarter, exceeding the 71.3% posted for fiscal year 2026.

A sober assessment of Nvidia's overall condition, however, suggests the market is probably right to impose this steep discount in response to Huang's departure. The company's extraordinary profitability stems from its virtual monopoly in the top segment of the GPU market. According to a senior industry executive whose company is an Nvidia customer, Nvidia is 18 to 24 months ahead of all rivals in the quality and performance of its leading chips. The complication, this person says, is Nvidia's dependence on Huang as both an unmatched innovator and an outstanding operator.

"Huang was the tech visionary and he also got Nvidia to operate incredibly well," the executive says. "The question is how much depended on Jensen, and how much of what he did can another person do?"

The executive emphasizes that Nvidia has maintained its huge margins by staying several steps ahead of the field without ever slowing down — one of the greatest feats in the annals of the technology industry. "That continuing advantage in chips is absolutely essential to Nvidia. It's an industry where you can't miss a step. If you do, others will catch up fast. Intel fell behind, and it never caught up."

Compounding the concern, competitors who have long complained about the high cost of Nvidia's chips are racing to build their own in-house alternatives — a group that includes AMD, Google, and Amazon. How quickly those alternatives close the gap, and whether Nvidia can hold its product cadence without Huang, are now among the clearest signals to watch.

It is also instructive to examine how Nvidia is deploying its vast free cash flow — the cash left after a company funds its operations and capital spending — to stay in front. Over the past four quarters, the company generated $105 billion in free cash flow. Historically, it has channeled a large share of that bounty into share buybacks, but in the 12 months ended in October, repurchases fell to just $32 billion. Nvidia's cash hoard of $100 billion is relatively modest and growing slowly.

So where is the remainder of that capital going? Nvidia has become well known for making large investments in its own customers, including SpaceX, CoreWeave, OpenAI, SoftBank's SB Energy, and Intel (for CPUs). It has also extended multibillion-dollar credit guarantees for projects built for hyperscalers and neoclouds where it supplies the GPUs. The practice of funding its own customer base is frequently questioned as "circular financing" — a label that reflects the loop in which Nvidia helps finance the very customers who buy its chips.

The open question is whether Huang felt obliged to pursue that course to protect Nvidia's future — locking in future sales as rivals hatch competing products at lower price points. In other words, Huang appears to have sought to forestall the commoditization of Nvidia, the threat that has eventually eroded the profitability of all tech hardware giants.

And that, ultimately, is the core problem. Huge prices and margins inevitably attract cheaper new supply. The market's judgment is that Nvidia's margins and profit growth will look far lower in the future than in the past — and that losing the genius who could hold commoditization at bay the longest was a big, big loss.

A former chief executive interviewed by Fortune took a contrarian stance, arguing that Nvidia's board made a serious mistake in not making every attempt to keep Huang. "I would have liked to see more facts, and maybe the board's got them," this person says. "The board was not thinking about shareholders. They were thinking about their reputations and not getting sued. They should have punished Huang and made it a very visible punishment like no equity or bonus for a couple of years. They should have found a different way to punish the guy that didn't damage shareholders."

The former executive concedes that this is probably a minority view. "People love to get on their high moral horse and that's what we're seeing," the observer says. "You can't just plug and play to replace someone so critical to the creation of a company's one-of-a-kind products and operation. A number of CEOs who couldn't say it publicly would probably say that the Nvidia board forgot what the role of a board should be, to protect shareholders, not themselves."

The breathtaking $2 trillion loss in market value may prove to be an enduring number — and not merely the result of the panic selling that most on Wall Street claim it to be.

This story was originally featured on Fortune.com.